Last month's London Underground bombings are a grim reminder of why the global market for homeland-security electronics is rapidly expanding. The tragedy also reminds us that more can be done to use advanced electronics to protect our citizens and their freedom of mobility.
Video surveillance recordings from the Underground have been key in identifying the perpetrators. Yet given the possibilities of artificial intelligence, it's compelling to consider how a video camera infrastructure could be put to more advanced work in identifying and preventing terrorist threats.
London's CCTV count is perhaps the most concentrated on the planet. Since the days of the IRA bombings, the city has installed CCTVs in big numbers. Over 10,000 cameras are in the central business district alone. With this density, average Londoners have their image captured as many as 300 times per day!
The promulgation of security cameras has raised some privacy concerns. The Institute for Applied Autonomy charts the locations of CCTV cameras and creates online guides to "the paths of least surveillance" for people who want to avoid being caught on camera. Yet cameras in subway stations are nearly universally accepted. People support the need for enhanced transport security, and as with going to an airport, they are willing to trade some personal freedoms for the necessities of public safety.
London's CCTV cameras record events, but for the most part they aren't monitored in real time for suspicious activity. The tedium and expense involved in watching 10,000 cameras on a typical day preclude effective manual monitoring. Here, AI could be applied to turn our cities' camera-intensive infrastructure into a more effective preventative tool.
Just down the pike from my office here in New Jersey, the Garden State Plaza is serving as an international testing ground for such a smart camera system. The system sends CCTV images to an AI software program that analyzes images for suspicious behavior—scanning for things like unattended bags, and people in heavy coats on hot days or loitering for an unusually long period of time. The New Jersey Institute of Technology is helping develop the system, which can filter the CCTV data and then alert security officers to manually check out any system-flagged activity. Results of the test will be announced by the end of the year.
While smarter CCTV solutions may be part of the security solution in public transport systems, trying to detect suspicious behaviors or apply face-recognition software in a densely packed subway station may prove to be the proverbial needle in a haystack. Still, the same close-quarters aspect of the transport system that makes it attractive to terrorists—secured entry and exit portals, closed passenger cars—also creates opportunities for new types of detection systems.
The subway stations in Washington, D.C., are the prototype for the Protect detection system. Designed in conjunction with Sandia National Labs, the D.C. system combines CCTV with sensors that sniff for hazardous chemicals. If toxins are detected, a red-alert signal is sent to subway operators who would immediately evacuate passengers. The Protect computer system maps out "hot zone" tunnels and stations that would likely be affected by the chemicals. It also routes passenger evacuations via the safe "cold zones."
Additionally, Sandia National Labs is working on an Explosives Detection Personnel Portal, a walk-through system for rapidly screening people for trace amounts of explosives. The portal forces air around a person to collect vapor and particles, then analyzes the chemicals using a mass spectrometer. The technology is in commercial use via Smiths Detection Sentinel in the CN Tower in Toronto.
Additional research in nanotechnology and MEMS sensors is dramatically changing the size and scope of sensing technologies as we move toward ubiquitous computing. Miniaturized, low-cost sensors can be coupled with wireless mesh networks to allow mobile, remote monitoring. Innalogic is a wireless software company specializing in software to communicate surveillance information wirelessly. The company worked with MSGI Security Solutions on the wireless network that supports remote, 24/7 bomb detection monitoring at the Statue of Liberty.
Hats off to all of you working on new technologies that can make our lives more secure.